This thesis project investigates navigational aspects of the upcoming JUICE (JUpiter ICy moons Explorer) mission, which is expected to deliver valuable data products for the scientific study of the Jovian system, and the three Galilean moons Europa, Ganymede and Callisto in particular. One such data product comes from the 3GM radio science payload, which is expected to contribute towards more accurate ephemerides of the Galilean moons and a refined study of their gravity fields. Accurate and reliable acquisition of these data depends greatly on the successful Guidance, Navigation and Control (GNC) operations, where an efficient GNC performance can unlock excess ∆V capabilities. It is the goal of this work to identify statistical ∆V savings from integration of in-mission radio science ephemeris improvements into the GNC operations. The resulting excess ∆V could be used to effectuate one of the mission enhancement and extension options (e.g. coverage of the Jupiter’s mid-high latitudes or a lower-altitude Ganymede orbit), the enablement of which ultimately motivates this thesis project. A navigational orbit determination (OD) framework was set up, modelled closely after the OD setup from the JUICE mission analysis team. An interface for simulating moon state knowledge updates from external ephemeris products was implemented. This setup was then extended to support a simplified statistical ∆V analysis. The external ephemeris products were modelled using an adopted high-fidelity OD setup with a coupled filter configuration, which simulates the uncertainty of in-mission ephemeris updates from the 3GM radio science data. These results were then used to simulate external ephemeris updates to the navigation OD and to evaluate their impact on the statistical ∆V expenditure. It was found that external moon ephemerides in general, and the simulated science data based OD solu- tion in particular, are not suitable for reducing the statistical ΔV cost of post-flyby correction manoeuvres. Within one to two encounters of a given flyby body, the nominal (update-free) navigation OD has improved moon state knowledge to a point where it no longer contributes significantly to size of the correction ma- noeuvres. To generate statistical ΔV savings, the moon state knowledge must be improved for the navigation of the mission’s first encounters. It was thus concluded that moon ephemeris improvements should not be provided from JUICE science data, but from observations that have been collected prior to mission. It was furthermore found that due to its extensive coverage, the navigation tracking data captures system informa- tion that the radio science data is less sensitive to. It could therefore be considered in the post-mission efforts to improve the high-accuracy ephemerides of the Galilean moons.

Owing to the assumed presence of sub-surface oceans, the Galilean satellites - Io, Europa, Ganymede, and Callisto - are among the most promising candidates for potential extraterrestrial habitats within our Solar System. To this end, the moons are going to be extensively studied by the upcoming JUICE and Europa Clipper missions. Ultimately, understanding the dynamics of the Jovian system is a means to shed light on the general existence and stability of these presumed habitable worlds as well as the formation and evolution of the entire Solar System. Yet, while the dynamics of Ganymede (in particular via the orbital phase of JUICE) and Europa (mainly via the various flybys of Europa Clipper) are going to be observed to an unprecedented level of accuracy, the absence of flybys of Io due to its harsh radiation-environment results in a significantly imbalanced data set. To stabilise the numerical data inversion, spaced-based imaging by the camera subsystem of JUICE is crucial to constrain the dynamics of Io. However, while the analysis of orbital dynamics is usually performed with respect to the centre-of-mass (COM) of natural satellites, optical space-based astrometry provides measurements of the position of a body's centre-of-figure (COF), introducing a discrepancy between the dynamical and observational model. Explicitly accounting for the offset between the observed centre-of-figure and the propagated centre-of-mass during ephemeris estimation, however, ensures the consistency of the dynamical and observational model. In turn, this allows us to assess the extent to which optical space-based astrometric observations might either validate the merely indirectly obtained radio science data or contribute to the overall orbital solution of Io. Finally, obtaining a measure of the offset between the centre-of-figure and centre-of-mass yields an entirely new constraint on the interior structure and composition of Io. In order to quantify the COF-COM-offset, we have simulated optical astrometric observations by JUICE and subsequently determined the formal uncertainties of the estimated offset using covariance analyses. Using suitably computed extit{a prioir} covariance matrices, we have constrained our analyses to the averaged propagated formal errors of Io that would arise from the radiometric tracking set-up of JUICE and Europa Clipper. We have found that the contribution of optical space-based astrometry to the COF-COM-offset of Io and its estimated state highly depends on the observations' quantity, quality, and geometry. Thus, an algorithm for the selection epochs at which images are to be simulated based on the main drivers - the absolute uncertainties and relative geometries of a series of observations - of the formal errors COF-COM-offset has been developed. However, owing to the largely equatorial alignment of JUICE with respect to Io - observations of the in-plane contribution have been found to be obstructed by the brightness of Jupiter. To maximise the scientific return of optical space-based astrometry, in particular, astrometry during the high-inclination phase has proven beneficial. Overall, significant constraints of the discrepancy between the centre-of-figure and centre-of-mass and the orbital solution of Io have been obtained. For an expectable number of about 1300~images being taken of Io, realistically attainable formal uncertainties in the estimated COF-COM-offset of no more than 300~metres have been obtained. Furthermore, since notable contributions to the orbital solution already occur for reasonable radio science true-to-formal-error ratios between two and five, we have concluded a high likelihood of space-based astrometry contributing to the orbital solution. This potential of space-based imaging to balance and contribute to the orbital solution of Io thus motivates future research concerning the offset between the centre-of-figure and centre-of-mass.